Macrophage-lymphocyte clusters in the immune response to soluble protein antigen in vitro. VII. Genetically restricted and nonrestricted physical interactions.

We have assessed the genetic restrictions on physical interactions between macrophages and central lymphocytes and between central and peripheral lymphocytes in antigen-specific macrophage-lymphocyte clusters with respect to I-region differences of inbred strains 2 and 13 guinea pigs. When using lymphocytes from guinea pigs immunized with DNP-OVA or DNP-GL in CFA, the antigen-specific interaction between central lymphocyte and macrophage requires that both cells be derived from animals syngeneic at the I-region of the major histocompatibility complex. In studies using antigens, the responses to which is under the control of MHC-linked Ir genes, macrophages from the responder, but not from the nonresponder parental strain support cluster formation with responder x nonresponder F1(2 X 13) T cells. In contrast, the physical interactions between central and peripheral T lymphocytes are not restricted by the I-region of the MHC and the peripheral lymphocyte need not be from an animal immune to the antigen used to drive macrophage central lymphocyte interactions.     

Induction of guinea pig antibody responses in vitro.

Guinea pig spleen cells cultured together with peritoneal exudate lymphocytes (PEL) were found to generate large numbers of antibody-forming cells (AFC) in vitro in response to hapten-protein antigens. Neither cell type cultured alone yielded appreciable responses. Strain 13 or F1 (Strain 2 X Strain 13) lymphocytes, but not those from strain 2 animals, are able to respond to the genetically controlled antigen, DNP-guinea pig albumin (DNP-GPA). Antisera directed against responder (strain 13) parent Ia antigens selectively blocked the generation of AFC by F1 (strain 2 X strain 13) spleen-PEL mixtures in response to DNP-GPA. Both allogeneic (strain 2) and syngeneic macrophages functioned equally well in presentation of DNP-GPA to strain 13 lymphocytes.     

Alveolar macrophages in pulmonary immune responses. I. Role in the initiation of primary immune responses and in the selective recruitment of T lymphocytes to the lung

Antigen inoculated intratracheally (IT) into animals can induce primary immune responses and selectively recruit specific T cells to the lung. In the current study, the role of alveolar macrophages (AM) in these two responses was investigated. Antigen-pulsed bronchoalveolar cells (BAC) inoculated IT into guinea pigs generated a population of immune T cells that proliferated in vitro on reexposure to antigen-pulsed macrophages (M?). The possibility that antigen-pulsed donor BAC shed antigen that was subsequently processed and presented by host M? was ruled out by genetic experiments. Thus, peritoneal exudate lymphocytes (PEL) from (2 X 13)F1 guinea pigs primed with antigen-pulsed BAC from strain 2 animals responded preferentially to antigen-pulsed strain 2 M? rather than to antigen-pulsed strain 13 M?. In a second set of studies, antigen-pulsed BAC inoculated IT into guinea pigs selectively recruited antigen-specific T cells to the lung. Genetic experiments verified that inoculated BAC were the source of the antigen-presenting cells responsible for selective recruitment. Thus, antigen-pulsed strain 2 BAC inoculated IT recruited a greater proportion of (2 X 13)F1 T cells that recognized antigen in the context of strain 2 M? than F1 T cells that recognized antigen on strain 13 M?. Taken together, these studies suggest that AM contribute to the regulation of pulmonary immunity by both inducing T lymphocyte immunity and selectively recruiting specific T cells to the lung.     

Generation of T helper cells in vitro. IV. F1 T helper cells primed with antigen-pulsed parental macrophages are genetically restricted in their antigen-specific helper activity.

A sequential culture technique for the in vitro induction and subsequent assay of T helper cells is employed to examine the histocompatibility requirements for antigen recognition by murine T helper cells. F1 T cells are primed in vitro with antigen-pulsed parental strain macrophages and tested for antigen-specific helper activity in cultures containing anti-Thy 1.2 serum and C treated spleen cells from hapten-primed parental or F1 mice. A semiallogenieic system is used and appropriate controls are included to avoid possible complicating effects of allogeneic interactions. The results indicate that F1 T helper cells preferentially stimulate carrier-specific anti-hapten plaque-forming cell responses in spleen cells which are H-2 identical with the macrophage used initially to prime the T cells. Parental spleen cell cultures do not respond to F1 T helper cells which were primed with the other parental strain macrophage. Supplementing this culture with macrophages which are histocompatible with those used to prime the F1T cells is sufficient to restore T helper cell activity. Thus, the genetic restriction described here is between the primed T cell and the macrophage used to elicit secondary responses and not between the T cell and B cell. The results in this semiallogeneic system, however, do not rule out the possibility of additional allogeneic genetic restrictions in the subsequent interaction of T cells with B cells.     

Nonrandom catabolism of proteins in the formation of antigenic peptide fragments

To examine the role of protein catabolism in the formation of antigenic peptide fragments, human fibrinopeptide-immune guinea pig T cells were stimulated with the large native molecule, human fibrinogen. Two different systems were tested. In the first, we determined responses by human fibrinopeptide B (hFPB)-immune T cells, to which strain (St.) 2 guinea pigs are responders and St. 13 are nonresponders, and by human fibrinopeptide A (hFPA)-immune T cells to which St. 13 are responders and St. 2 are nonresponders. Of interest in this comparison is that both hFPA and hFPB are amino terminal peptides on the A and B chain of fibrinogen, respectively, and are readily cleaved by thrombin during fibrin formation and by other trypsin-like enzymes, leaving a carboxyl terminal Arg. Thus, if fibrinogen catabolism occurred, both antigenic peptides should be equally represented for availability in T cell responses. It was found that hFPB-immune St. 2 T cells responded to fibrinogen, but no response was observed with hPFA-immune St. 13 T cells cultured with fibrinogen. To rule out that there was a general catabolic defect in St. 13 antigen-presenting cells, fibrinogen was presented by (2 X 13)F1 macrophages to fibrinopeptide-immune parental T cells. Again it was found that F1 macrophages could present fibrinogen to hFPB-immune T cells but failed to present hFPA. In another comparison, responses with fibrinogen were also determined with des-ARg-hFPB, which lacks the carboxyl terminal Arg of hFPB, to which St. 13 are responders and St. 2 are nonresponders. The advantage of this comparison is that both antigenic determinants are contained within the same small peptide. St. 13 des-Arg-hFPB-immune T cells failed to respond in vitro by culture with human fibrinogen, suggesting that these antigenic determinants are not produced from larger peptides or proteins containing those determinants. To rule out the possibility that this was only an in vitro phenomenon, guinea pigs were immunized with the larger protein, the B chain of fibrinogen, and the immune T cells were examined for responses to fibrinopeptides derived from the B chain. Immune St. 2 T cells responded to hFPB but not to des-Arg-hFPB, whereas St. 13 T cells remained unresponsive with both peptides. These results indicate that proteolysis of larger proteins to form small antigenic peptides is not a random event and that not all potential antigenic determinants contained in a protein are produced during antigen processing.     

Nature of the antigenic complex recognized by T lymphocytes. V. Genetic predisposition of independent F1 T cell subpopulations responsive to antigen-pulsed parental macrophages.

Nonimmune (2 x 13)F1 guinea pig T lymphocytes initially stimulated with trinitrophenyl- (TNP) modified macrophages from one parental strain and then treated with bromodeoxyuridrine (BUdR) and light are unable to be primed subsequently with TNP-modified macrophages of the same parental strain. In contrast, these BUdR and light treated F1 T cells can be primed with TNP-modified macrophages of the other parental strain. These results demonstrate that in a nonimmune F1 animal two T cell subpopulations exist before priming that are genetically predisposed to respond to antigen associated with macrophages derived from one or the other parent.     

The syngeneic mixed leukocyte reaction: the genetic requirements for the recognition of self resemble the requirements for the recognition of antigen in association with self

We have used cells from inbred strain 2 and strain 13 guinea pigs in order to define further the role of Ia antigens in the syngeneic mixed leukocyte reaction (MLR). The guinea pig syngeneic MLR resembled the autologous MLR in man in that it demonstrated both memory and specificity. The Ia antigens appeared to be the proliferative stimuli in that the primary stimulator cell was an Ia-positive adherent peritoneal exudate cell (PEC) and the reaction could be specifically inhibited by anti-Ia sera directed to the stimulator cell. We also demonstrated the existence of two (2 x 13)F1 T cell populations that were capable of reacting to one or the other parental PEC in the absence of any known exogenous antigen. These results suggest that the syngeneic MLR may represent T cell activation mediated through a receptor for self Ia.     

Failure of ovalbumin associated with allogeneic spleen ceels to stimulate proliferation of lymph node cells of immune mice.

Ovalbumin-pulsed spleen cells were found to stimulate thymidine uptake of lymph node cells of syngeneic mice immunized with ovalbumin in complete Freund's adjuvant after treatment of spleen cells with Mitomycin C but not after heating the spleen cells at 56degrees for 30 min. Ovalbumin-pulsed spleen cells of allogeneic mice failed to stimulate the immune lymph node cells more than unpulsed cells, although a net increase in the thymidine uptake above the allogeneic stimulation was observed when free ovalbumin was added to the mixed culture. To eliminate the high background of the mixed lymphocyte reaction, F1 mice were made chimeric with bone marrow of one of the parental strains. Using lymph node cells of the immunized chimeras, the stimulation by pulsed spleen cells was much greater when antigen was presented on cells of the parental strain used for bone marrow injection than when presented on cells of the other parental strain.     

The role of B lymphocytes in cell-mediated immunity II. Delayed hypersensitivity induced by dinitrophenyl-ficoll in dinitrophenyl-keyhole limpet hemocyanin-immunized guinea pigs.

Dinitrophenyl (DNP)-Ficoll will elicit typical delayed hypersensitivity skin reactions in guinea pigs immunized with DNP-keyhole limpet hemocyanin (KLH). We observed that lymph node cells (LNC) from these animals produced the lymphokine, monocyte chemotactic factor (MNL CTX) when stimulated by DNP-Ficoll in vitro. This response was antigen and hapten specific since LNC from nonimmune guinea pigs or those immunized with nonDNP containing antigens were not stimulated by DNP-Ficoll. Lymph node cells were fractionated into T- and B-cell-enriched populations to determine the nature of the DNP-Ficoll-responsive cell. Only the B-lymphocyte-enriched population produced MNL CTX in response to DNP-Ficoll. The purity of the B-cell population was demonstrated by its failure to respond to PHA and by the fact that B cells derived from DNP-although they could no longer respond without T-cell help to the T-dependent antigen, DNP-OVA. These findings suggest that the hapten-specific response of guinea pigs to DNP-Ficoll may be a form of B-cell-mediated delayed hypersensitivity.     

The role of myelin lipids in experimental allergic encephalomyelitis. III. Transfer of suppression from guinea pigs recovering from EAE, induced by myelin basic protein--galactocerebroside complexes

Juvenile strain 13 guinea pigs were immunized with myelin basic protein (MBP) combined with galactocerebrosides (MBP + GC) or with total myelin lipids without GC [MBP + (TL-GC)] in CFA. Control animals received dinitrophenylated-ovalbumin (DNP-OA) in CFA, CFA or IFA alone. The animals injected with MBP + GC showed a higher rate of recovery from the first EAE episode (83%) than those treated with MBP + (TL-GC) (50%). With the exception of the group treated with IFA alone, all animals were refractory to EAE following rechallenge with MBP in CFA 90 days after the first exposure. The in vitro proliferative response to MBP, of peripheral blood lymphocytes (PBLs) derived from guinea pigs freshly sensitized to MBP in CFA, was drastically suppressed in the presence of PBLs from animals injected with MBP + GC. Upon transfer to normal syngeneic recipients, spleen cells from MBP + GC-treated animals completely suppressed the clinical and histological manifestations of EAE following recipient challenge with MBP in CFA. Cell-free supernatants from PBLs and spleen cells of strain 13 guinea pigs treated with MBP + GC inhibited lymphocyte proliferation to MBP, of allogeneic responder cells, and spleen cell supernatants completely suppressed the induction of EAE upon transfer to allogeneic recipients. Suppression could not be transferred with cells from other treated groups. These results suggest that animals immunized with MBP + galactocerebrosides in CFA develop suppressor cells that may be in part responsible for the recovery from the first EAE episode and for protection against rechallenge with MBP in CFA. Their cell-free supernatants act in an MHC-nonrestricted fashion. These results do not rule out an additional protective mechanism since all animals exposed to CFA were refractory to rechallenge despite lack of demonstrable suppressor cell activity.     

Suppressing effects of purified eosinophils derived from guinea pigs sensitized with Ascaris antigen on lymphocyte-blastformation

Highly purified eosinophil (EOS) fractions were obtained from peritoneal exudate cells of guinea pigs immunized with with Ascaris lumbricoides suum antigen (Asc). These Eos suppressed the in vitro DNA synthesis of the lymph node cells (LNC) sensitized with Asc and then activated by this antigen or phytohemagglutinin (PHA). Although addition of Eos did not effect the viability of LNC in vitro, the blastformation of LNC was suppressed remarkably when 5-10 X 10(5) purified Eos were added to 10(6) LNC within 48 hr after the start of stimulation by Asc. The suppressive effects of Eos on the blastformation of LNC immunized with complete Freund's adjuvant with or without ovalbumin were observed when stimulated with purified protein derivates or ovalbumin. Such suppression were observed beyond the barrier of animal strain specificity; Eos from Hartley guinea pigs suppressed proliferation of LNC from either strain 13 or strain 2, and Eos from strain 13 suppressed that from strain 2. Such suppressing activity of Eos was reduced by heating them at 56 C for 1 hr or by sonication.     

Immune response gene control of antibody specificity

The expression of the histocompatibility-linked PLL Ir gene was investigated in guinea pig B cells. Strain 2 and F₁ (2 × 13) guinea pigs, immunized with the αDnp-Lys₉, produce both T cells and antibody which are equally discriminatory for αDnp-Lys₉. In contrast strain 13 (PLL Ir gene negative) guinea pigs immunized with αDnp-Lys₉ do not develop specific T-cell responses and the antibody produced while restricted in heterogeneity cannot differentiate the immunizing antigen from Dnp-OH. However, if in a F₁ (2 × 13) environment, PLL Ir gene-negative B cells are provided with F₁ (2 × 13) T cells they express the ability to make antibody as specific and discriminatory as the antibody produced by PLL Ir gene-positive B cells. These findings strongly suggest that in the guinea pigs the PLL Ir gene defect is localized to the T cells and that the repertoire of specificity of B cells is similar if not identical in both responder and nonresponder animals. In addition these observations support the notion that the cellular locus for the PLL Ir gene expression in the guinea pigs is limited to T cells and not to macrophages and B lymphocytes.     

Specific absorption of T lymphocytes committed to soluble protein antigens by incubation on antigen-pulsed macrophage monolayers.

Monolayers of macrophages (Mphi) pulsed with antigen were used as immunosorbents for T lymphocytes from guinea pigs primed to soluble protein antigens. T lymphocytes were cultured on the Mphi monolayers for 4 hr, then aspirated and reincubated on a fresh monolayer pulsed with the same antigen for a second and a third step. T lymphocytes so treated were selectively deprived of cells responding in assay for antigen-dependent proliferation against the antigen used for pulsing the absorbing monolayer, but maintained their response to other antigens. The lymphocytes adhering to the Mphi of the absorbing monolayer were capable of giving a full response to the antigen used for pulsing the Mphi of the monolyers. The proliferative response of F1 T lymphocytes to antigen in association with Mphi of either parental strain could be absorbed leaving the response to antigen in association with Mphi of the other parental strain. The absorption of the proliferative response was not inhibited by addition of excess soluble antigen to the medium of the absorption culture. Our results indicate that specific guinea pig T lymphocytes responding by proliferation to soluble protein antigens recognize and bind specifically to a complex of Ia antigen and protein antigen at the surface of the Mphi.     

Macrophage activation of allogeneic lymphocyte proliferation in the guinea pig mixed leukocyte culture.

The role of the macrophage in the guinea pig mixed leukocyte culture was investigated. Macrophages obtained from oil-induced peritoneal exudates, peritoneal wash-out cells, spleen, and alveolar washings were found to be effective stimulators of allogeneic lymph node and splenic lymphocyte DNA synthesis. The stimulatory properties of macrophages proved radioresistant but viability dependent. Unfractionated lymph node cells or adherence column purified lymph node lymphocytes and thymocytes were only minimally active as stimulators, even in the presence of macrophages syngeneic to the responder lymphocytes. Allogeneic fibroblasts, polymorphonuclear leukocytes, L2C leukemia cells, and xenogeneic (murine) macrophages failed to simulate. These data provide evidence that the macrophage is the predominant stimulator of the mixed leukocyte culture in the guinea pig.     

Genetic restrictions in the development of antibody responses to L-glutamic acid60-L-alanine30-L-tyrosine10 by nude mice implanted with semiallogeneic thymus glands

Athymic nude mice implanted with F1 thymus glands were used to investigate genetic restrictions regulating T cell-macrophage (M phi) interactions in the development of antibody responses to GAT. Spleen cells from conventional mice developed comparable primary plaque-forming cell (PFC) responses when stimulated by syngeneic and allogeneic GAT-M phi. However, spleen cells from strain A nude mice implanted with (A X B)F1 thymus glands were tolerant of strain B alloantigens and developed GAT-specific PFC responses to strain A GAT-M phi and allogeneic strain C GAT-M phi, but failed to respond to strain B GAT-M phi. The lack of primary GAT-specific PFC responses by spleen cells from (A X B)thy----A nude mice stimulated by strain B GAT-M phi was not due to detectable suppressor mechanisms. However, an allogeneic effect stimulated by H-2- or non-H-2-disparate GAT-pulsed or unpulsed M phi was able to overcome the inability of spleen cells from (A X B)F1 thy----A nude mice to respond to strain B GAT-M phi. Furthermore, the inability to respond to strain B GAT-M phi was overcome by the addition of supernatant fluids from independent cultures of H-2-disparate cells. These results 1) demonstrate that T cells from A nude mice implanted with (A X B)F1 thymus glands did not recognize nominal antigen in the context of B MHC antigens, and 2) suggested that the T cell repertoire was altered in strain A nude mice implanted with (A X B)F1 thymus glands, such that T cells that could recognize GAT in association with strain B MHC antigens were functionally deleted.     

Successful immunization against experimental allergic encephalomyelitis with myelin basic protein-sensitized allogeneic lymphocytes

Prevention and suppression of experimental allergic encephalomyelitis were demonstrated in rats, guinea pigs, and rabbits immunized with allogeneic, but not with syngeneic lymphocytes from susceptible donors sensitized to myelin basic protein (MBP). Donor lymphnode, splenic, or peripheral blood lymphocytes were effective in inducing a state of unresponsiveness to an encephalitogenic challenge in either of the three species. Unresponsiveness was not obtained in recipients immunized with sensitized allogenic lymphocytes and simulatenously challenged with MBP suggesting that a time lapse between immunization and challenge is necessary for the development of protective immunity. Induced in immunized recipients, unresponsiveness was transferred into normal syngeneic recipients with immunoglobulin-G (IgG) isolated from protected donors before challenge. Furthermore, both immunized and IgG recipients failed to develop cell-mediated immunity after challenge with MBP. The results show that prevention and suppression of EAE was mediated by antibodies which inhibited the development of delayed type hypersensitivity to the challenging antigen.     

Augmentation of syngeneic tumor-specific immunity by semiallogeneic cell hybrids

Hybrid cell lines were established from fusions between lipopolysaccharide- (LPS) stimulated C57BL/6J spleen cells and MPC-11 tumor cells (45.6TG1.7, abbreviated M45), and were tested for their ability to immunize semiallogeneic mice against a parental tumor challenge. These hybrids were tumorigenic in syngeneic (BALB/c X C57BL/6J) F1 (CB6F1) mice but did not grow in semiallogeneic (BALB/c X A/J) F1 (CAF1) mice. All hybrids express both parental major histocompatibility antigens (H-2b and H-2d) as detected by indirect immunofluorescence and by their ability to function as either stimulators or targets for allogeneic cytotoxic lymphocytes (CTL). M45 tumor-associated antigens (TAA) were expressed on the hybrid surface as shown by their ability to act as either stimulators or targets for syngeneic CTL specific for M45 TAA. Immunization of semiallogeneic CAF1 mice with the hybrids i.p. followed by a challenge with M45 tumor cells resulted in extended survival when compared to untreated mice or animals immunized i.p. with M45 tumor cells. This immunity was specific and was not due to an allogeneic effect; immunization with an unrelated H-2bd tumor, 70Z/3, or H-2bd B6D2F1 spleen cells or with semiallogeneic spleen cells plus M45 did not protect mice from M45 challenge. Interestingly, prophylactic priming with semiallogeneic hybrid tumor cells or parental myeloma cells led to M45-specific CTL and "help" for an in vitro CTL response; however, the degree of CTL priming by hybrid tumors was not augmented when compared to the level of CTL achieved with parental tumor alone. Hence, stimulation of CTL activity per se by hybrid tumor cells cannot explain the protective effect of hybrid tumor immunization. These studies nevertheless confirm that semiallogeneic hybrids, which we show express TAA and alloantigens, can be used to immunize mice against a lethal syngeneic myeloma tumor challenge.     

Specific binding of T lymphocytes to macrophages. V. The role of Ia antigens on Mphi in the binding.

The effect of anti-Ia alloantiserum on the capacity of selected peritoneal exudate lymphocytes (selected PEL) to bind to antigen-pulsed F1 (responder x nonresponder) macrophages was investigated. With the use of selected PEL for antigens under Ir gene control, it was shown that anti-Ia serum to the responder haplotype blocked adherence of selected PEL to antigen-pulsed macrophages whereas anti-Ia serum to the nonresponder haplotype did not. The target cell of the anti-Ia alloantiserum appeared to be the macrophage because anti-13 Ia in contrast to anti-2 Ia did not inhibit binding of F1 (2 x 13) DNP-GL selected PEL to DNP-GL pulsed strain-2 Mphi (responder strain). Taken together with previous experiments that indicate that an antibody to the native protein antigen employed is unable to block specific binding, the present results suggest that T cells may recognize fragments of exogenous antigen in association with Ia molecules.     

Responder strain-specific enhancement of endothelial and mononuclear cell Ia in delayed hypersensitivity reactions in (strain 2 X strain 13)F1 guinea pigs

This study was undertaken to test the hypothesis that preferential responder strain-specific Ia expression can be detected in delayed hypersensitivity (DH) skin reactions. Seven adult (strain 2 X strain 13)F1 and two strain 13 guinea pigs were sensitized with poly-L glutamic acid-lysine (GL), poly-L glutamic acid-tyrosine (GT), and bovine insulin in complete Freund's adjuvant, and were skin tested with GL, GT, PPD, bovine insulin, porcine insulin (which has the same B chain as bovine insulin), and saline. Strain 2 guinea pigs react with bovine insulin A chain, GL, and PPD but not with GT or the bovine insulin B chain, whereas strain 13 guinea pigs react with bovine insulin B chain, GT, and PPD but not with GL or bovine insulin A chain. The (2 X 13)F1 animals had positive DH responses to GT, GL, PPD, and bovine insulin. At 24 hr, areas of induration were measured and the test sites and draining lymph nodes were biopsied. Cryostat sections were stained with monoclonal antibodies to strain 2 Ia, strain 13 Ia, and Ia framework determinants with immunoperoxidase. Stained dermal and subdermal inflammatory cells and vessels were counted on coded slides. In GT tests, there was more staining of dermal and subdermal cells and vessels for strain 13 Ia than strain 2 Ia (p less than 0.02). In bovine insulin tests there was more staining of dermal cells and vessels for strain 13 than strain 2 Ia (p less than 0.05). In GL tests there was more staining on dermal vessels and subdermal cells and vessels of strain 2 Ia than strain 13 Ia (p less than 0.05). There was much greater staining of strain 2 Ia of dermal cells and vessels in GL tests compared with strain 2 Ia staining in GT and bovine insulin tests (p less than 0.02, cells; p less than 0.01, vessels). No significant differences between strain 2 and strain 13 Ia expression were found in PPD, porcine insulin tests, saline controls, or in lymph nodes that drained sensitization sites from animals in which GL and GT had been injected on different sides. Anti-Ia framework expression generally correlated with the greater parental strain Ia in each reaction. These findings and previous observations in experimental allergic encephalomyelitis suggest that responder type Ia may be selectively found in vivo on mononuclear and endothelial cells in sites of T cell-mediated hypersensitivity reactions.     

The specificity of delayed hypersensitivity to ABA-tyr-pulsed and ABA-coupled macrophages

Guinea pigs immunized with ABA-tyr in CFA respond well by skin test to ABA-tyr-pulsed macrophages but poorly to ABA-coupled macrophages and not at all to ABA-coupled red cells, thymocytes, or L₂C cells. On the other hand, guinea pigs immunized with ABA-coupled macrophages do not respond to ABA-insulin or ABA-tyr-pulsed macrophages but do respond to ABA-coupled macrophages. Similarly guinea pigs immunized with Ars-NCS-coupled macrophages respond only to the homologous antigen. The specificity of these reactions is determined by how ABA is associated with the Ia-positive accessory cell. The presence of Ia molecule is not a sufficient condition since neither Ia-positive L₂C cells nor spleen cells depleted of adherent macrophages are effective as immunogens or elicitors of response when coupled with ABA. These results suggest that the topography of the ABA and Ia complex formed on the accessory cell is the prime determinant of specificity for T-cells responses.